IC/85/235 INTERNATIONAL CENTRE FOR THEORETICAL PHYSICS SOME CRITICAL CONSIDERATIONS ON THE PRESENT EPISTEMOLOGICAL AND SCIENTIFIC DEBATE ON QUANTUM MECHANICS GianCarlo Ghirardi INTERNATIONAL ATOMIC ENERGY AGENCY UNITED NATIONS EDUCATIONAL SCIENTIFIC AND CULTURAL ORGANIZATION 1985 MIRAMARE TRIESTE 1. INTRODUCTORY COM3THERATIONS Tnlf^national Atomic Energy Agency and In view of the fact that the debate on the epistemolopi^-;! ."'.nd conceptual implications cf quantum theory is often vitiated, in our iiitions Educational L ::ienti f"ic and Cultural Organization opinion, by gross misunderstradings, we consider it appropriate to focus on some methodological principles which must be kept in mind in order to guarantee the necessary level of rigor of such a discussion. IITERNATlODlAL CEHTHE FOR THEORETICAL PHYSICS The occurrence of misunderstandings on this matter is not surprising if one takes into account the conceptual difficulties that the formalism presents, on which the critical debate is still alive, and which have not yet been fully clarified. One can mention that, already in the early stages of the development of the theory, even eminent scientists SOME ClilTICAL CONSIDERATIONS OH THE PRESENT EPI31EM0L0GICAL have been unable to fully appreciate^ the crucial points and the subtle AND SCIENTIFIC DEBATE ON QUANTUM MECHANICS * implications of the deep analysis of Einstein-Podolsky and Rosen^(EPR). We want to stress strongly the necessity of logical rigor and clarity in carrying on the debate. This can be guaranteed only if the terms that are used and the conceptual framework in which each critical investigation finds Its place, are very precisely defined. Obscurities, lack of precision, and rash extrapolations, seriously damage this kind GianCarlo Ghirardi of research, and, in particular, they reinforce the wrong and naive, but very dangerous conviction, shared by several members of the International Centre for Theoretical Physics, Trieste, Italy scientific community, that to investigate the conceptual and and epistemological implications of the theory is a more or less useless Dipartimento di Fisica Teorica, University di Trieste, Italy. task. This, in turn, increases the division between the so called "two cultures," with all its harmful consequences. To avoid the above misunderstandings, it is first of all necessary that the object of any critical analysis be defined precisely, in particular by making absolutely clear whether the analysis involves the ABSTRACT direct consequences of the formal scheme or only one of its possible interpretations. The meaning of the terms that are used must be made Some general methodological considerations aimed to guarantee the very precise; e.g., a source of misunderstandings is often the use of terms such as "realism," tout court, to describe both the very necessary logical rigor to the present debate on quantum mechanics are precise requirements of reality put forward by EPR and the much more presented. In particular some misunderstandings about the implications stringent requirements coming from various forms of philosophical realism. The same applies to the term "locality," which has very of the critical analysis put forward by Einstein Podolsky and Eosen (EPP.) different conceptual connotations in the critical investigations of EPR which can be found in the literature, are discussed. These misunder- and those of J.Bell. Analogous difficulties arise when it is claimed that a "new" interpretation of the formalism resolves all its standings are shown to arise from possible underestimates, overestimates conceptual difficulties, without confronting the proposed interpretation and misinterpretations of the EPR argument. It is argued that the with all the known crucial points of the theory. As an example, we can mention that it is sometimes stated that alternative interpretations, difficulties pointed out by EPR are, in a sense that will be defined such as the "propensity interpretation" or the Introduction of non- precisely, unavoidable. A model which tries to solve the difficulties KoJmogoroffian models of probability, solve the EPR paradox, without facing the problem of identifying which "elements of physical reality" arising from quantum non separability effects when macroscopic systems cs.r. be attributed to the physical system under investigation within the are involved, is briefly sketched, proposed interpretation, a point which is central for discussing the EPR reasoning. Let us now specify some general methodological points. MIRAMABE - TRIESTE September 1985 -2- * Invited talk at the Meeting on ''The Paradox or.1 Physical Reality", 11-13 April 1985, Cesena, Italy. 2. GENERAL METHODOLOGICAL COPTS'. r'FRAVIOK',; This theorem implies that different (as operators) states involve It is useful to recall'-' her? ;,..••:;; posaitli? apnr'.-:r-hes to n critical different physical predictions and are thus, In principle, experimentally analysis of an established theoretical ycherr-e, distinguishable. This in turn implies that one cannot describe the same A. One can check the compatibility of the theory with general physical situation with different statistical operators. I shall not ideas abouc physical reality and the level of knowledge we can get of enter the debate on this assumption (which could be questioned in some it. Thic kind of analysis is,in our opinion, not only legitimate, but necess.T . As a typical example of this line of approach, we can refer cases) since I will bear in mind, in discussing the EPR paradox, such to the -B paper. simple systems as the spin states of a two-particle system, in which B. One can check the compatibility of the theory with other case the assumption is certainly correct. physicil principles, which are viewed as "true" by the scientific There is a simple theorem^) of quantum mechanics which is extremely community, even though they are not included among t!v= axioms of the important for an appropriate understanding of the requirement of theory considered- Into such a context one can fit aJ5 recent attempts^ completeness of the theory, even though it is not often quoted in the to prc/e that the quantum formalism would allow, through! wave packet discussions on the EPR paradox. reduction, faster—than-light communication between different observers. Theorem 2. For an ensemble of composite systems S = S^+ §2' wlrlich As we will show, this line cannot leau to any significant result. is a pure state, the necessary and sufficient condition that there will C. One can check the internal consistency of the theory. Within exist a complete set of commuting observables belonging to one of the this approach fall all investigations direct to clarify whether one can two subsystems S^ and Sg , for which the probability of getting a describe the functioning of the measuring apparata, as postulated in given set of results equals 1, is that the state of the system be the quantum theory, in terms of the dynamics of the microconstituents of projection operator on a factorized state vector these objects. This crucial point of the quantum theory of measurement has in our opinion not yet been completely clarified, p = (5) Remarks on Theorem 2: We shall call states of type (5) 3. THE EPR PARADOX "factorized states" and denote them by p . In view of this theorem, it is legitimate to think of an ensemble as composed of subensembles, such In spite of the many investigations on the EPR paradox which can that for each subensemble the result of a measurement of an appropriate be found in the literature, we consider it appropriate to reconsider it observable of subsystems S\ or S2 can be predicted with certainty if in a coincise way, and to focus on some points which will be useful in and only if p is a statistical mixture of factorized states illustrating some misunderstandings about this subject. Let us recall, first of all, some basic points of the formalism. P = t PiPF1 In quantum mechanics an ensemble is described by the statistical operator p, which will be briefly called "a state" and is a trace-class Note also that, in the case of a spin - Y2 particle, any spin component operator with trace l. We will assume that no superselection rule is is by itself a complete set of commuting observables. present for the systems of the ensemble. We can now discuss in detail the EPR argument. We will use a Rule 1. The physical predictions of the theory are obtained version of it which is slightly modified with respect to the original according to the following scheme. Let A be an observable, and P^ the one, but is now currently used. H/e will summarize the argument in a spectral family of projection operators associated with the eigenvalues sequence of steps denoted by capital letters from A to E. ak of A. If the ensemble is in the state p, the probability P(A=ak|p ) A. Einstein's reality requirement: If, without in any way of finding the result ak in a measurement of A, is given by disturbing a system, we can predict with certainty the result of a measurement of one of its observables, then there exists an element of P(A = ak|p) - Ir Pk p (1) physical reality associated with this prediction. Comments on A, On the basis of Theorem 2, we can state from which it follows that the mean value of the results obtained in a {a) If measurement of A is <A > = Tr A p (2) then we can attribute elements of physical reality to the individual Rule 2. Reduction of the wave packet: For an ensemble in the state subsystems of the composite systems. PD • a non selective measurement of an observable A induces the (b) In the quantum formalism, the formal counterpart of the existence following change of the statistical operator of elements of physical reality (possibly different from subsystem to subsystem) for all individual subsystems of the composite systems is (3) that the ensemble can be described by a statistical mixture of factorized states, Pa and p.